Bing Predicts: Making Sense of NCAA® March Madness®

With the arrival of spring, everyone knows it is time for warmer weather, blooming flowers, and beating your friends with your March Madness® brackets. While we can’t help with picnic plans or allergy relief, Bing Predicts returns again to provide forecasts and insights into this year’s NCAA® tournament for both men’s and women’s draws to help you fill out your bracket. Click here for the full smarter bracket experience so you can get an edge on the madness.

On the men’s side, by multiple metrics, the four No. 1 seeds are as up-for-grabs as any ‘top four’ since the tournament expanded to 64-plus teams. The combined 23 losses are the most by any group of No. 1 seeds going into the tournament, as compared to 12 in 2014 and nine in 2015 and 20 in 2000, the previous high. The greater amount of parity makes the statistical insights we provide all the more salient for teams you may not be as familiar with.

For the women, there is a clear overall No. 1 in Connecticut. This team seeks to win their 4th straight National Championship. They enter the tourney in great shape as the lone undefeated team in college basketball and appear poised to defend their title yet again.

Last year, Bing Predicts was in the upper 30 percent of all national brackets, beating those published by Google, Facebook, and Sports Illustrated. We were one of just a few who had eventual winner Duke listed as the 2nd most likely winner—most everyone had Kentucky incorrectly as the predicted winner due to their dominant regular season at the top.

Men’s bracket:

The madness expected every March started early as only ten of the 31 top seeds, earned by being regular season champions, won their conference tournaments. Three of the ten came from major conferences which resulted each securing a No. 1 seed: North Carolina (ACC), Kansas (Big 12), and Oregon (Pac 12), leaving effectively only seven teams from the small conferences which won both. What this actually means is that there may be less uncertainty in the early rounds as the 13 to 16 seeds coming in are already underdogs which have pulled off surprises to get their way into the tournament. A rough analogy would be this year’s pro football playoffs. Ignoring how the football seeds are set, recall in the opening weekend that all of the road teams won. Not surprisingly, all of the home teams defeated these road teams in the second weekend. Because of this, we don’t see any of the top four seeds (top 16 teams) losing in the first round this year, but this doesn’t mean there won’t be a handful of upsets predicted.


In particular, our models predict that a couple of the teams which won both their regular season and conference tournaments will indeed pull off surprises.

First, Thursday afternoon, we predict Arkansas-Little Rock will pull off the first of the tourney’s upsets with a win over No. 5 Purdue. At 29-4, only overall No.1 Kansas has a better record, giving them a profile akin to the 2014 No. 12 seeded Stephen F. Austin team which entered the tournament 31-2 – namely, a consistent season playing mid-major teams, a regular season conference title, and a conference tournament title. We also think they’ll do one better by also reaching the Sweet 16®.

Next, on Thursday evening, we see 29-5 UT-Chattanooga, a No. 12 seed with road wins over No. 7 seed Dayton and bubble team Georgia this season, pulling off an upset, beating Indiana in Des Moines, IA where there might not be a crowd advantage for the Hoosiers playing in Big Ten rival Iowa’s home state. The Southern champion Mocs will try to emulate their 1997 team which, as a No. 14 seed, upset No. 3 Georgia.

Our models then see Gonzaga upsetting No. 6 seed Seton Hall in one of the late games Thursday. Gonzaga has the 3rd longest active tournament streak at 18 appearances while Seton Hall makes their first appearance in ten years. The Pirates finished strong to win the Big East tournament title but have a Jekyll and Hyde resume, beating Xavier twice and Villanova while losing to Long Beach State and at home to Creighton. Our models also think Gonzaga is under-rated as a No. 11 seed, perhaps partially due to several close losses (six of their seven losses were 5 points or less).

Of the lower seeded teams, we also predict that No. 10 seeds Pittsburgh and VCU will pull off wins as well as No. 9 seed Butler. These may be considered upsets but historically 9 and 10 seeds have won their fair share of games versus the 8 and 7 seeds, so these may not be as shocking.

If you’re looking for what we expect to be the closest game for a top 4 seed, it’s Cal vs Hawaii. The models have Big West champions Hawaii playing a close contest against No. 4 Cal. Interestingly, 27-5 Hawaii played all but one of their games in the states of Hawaii and California. Playing in Spokane, WA keeps them on the West coast to minimize their travel, and being banned from next year’s tournament gives them extra motivation to do well this year. 23-10 Cal didn’t have strong road wins on their resume, with a road win at 18-14 Washington being their best win. The 11am PT start (8am Hawaii time) is the earliest start for them this year which hinders their chances and they also don’t have any big road wins, so we have Cal just barely winning.

NCAA® Final Four® predictions:

Most bracket systems employ exponential scoring, meaning that four correct picks in the first round equals two in the second and one in the Sweet 16®. So a key to doing well is making sure you don’t eliminate the teams who will go far, as picking a great upset here and there is nice but one missed Final Four® team in the early round and that quadrant of your bracket is busted.

No NCAA champion has ever suffered a blowout loss (30+ points) in the regular season. Xavier has had a solid season and we think deserving of a high seed, but they did have a 31-point loss to Villanova early this season and thus might not have the consistency to put together a streak of 6 wins to win the tournament. The same goes for Utah which lost the Pac-12 title game by 31 points. Miami’s fate also doesn’t bode well with a 25-point loss at North Carolina in late February as well as Dayton which lost by 29 to Xavier.

The Final Four Bing predictions include the #1 team this year (Kansas), the defending champion (Duke), and two other teams with strong tournament pedigrees (Michigan State and North Carolina).

Kansas, winners of 14 in a row, is expected to win four more to get to the Final Four. As one of the hottest teams in college basketball right now, Kansas might make its way to Houston without any major challenges despite what many say is a tough South region.

Duke, the defending champions, have quality wins over Virginia (albeit controversial) and at North Carolina, comparable to the top wins Oregon has over Utah and Arizona which are coupled by losses to Boise St, UNLV, and Stanford. In a close contest, we have Duke upsetting Oregon and then beating Oklahoma to make the Final Four.

Michigan State, No. 1 early in the season, is led by Big Ten player of the year Denzel Valentine. With Tom Izzo’s experienced coaching they appear poised to reach yet another Final Four. They are matched up in the same region as Virginia for the 3rd straight year (Michigan State won as a No. 7 seed over No. 2 last year and as a No. 4 seed over a No. 1 the year before)

North Carolina, a team which won both the ACC regular season and post-season, enters with quality wins over Virginia, Miami, Duke, and Maryland. Their six losses were by only a combined 22 points. We see them being the fourth team in the Final Four.

Our models predict Kansas beating North Carolina in the finals on Monday, April 4.

Women’s bracket:

The announcement of the women’s bracket won’t be until Monday evening (7pm ET), so the link above will have our predictions and the field only after that time, but we have some early thoughts on who the No. 1 seeds will be and who might make deep runs in the tournament.

No. 1 seeds:

The top 4 teams in women’s basketball finished an amazing 125-3 (compare this to 112-23 for the four No. 1 men’s seeds) with two of those losses coming from Connecticut wins over Notre Dame and South Carolina. Excluding games between these four teams, they were 123-1, with the lone loss Baylor’s road defeat by Oklahoma State. We project that these four teams, Connecticut, Notre Dame, South Carolina, and Baylor, will get the four No. 1 seeds when announced tomorrow night.

Final Four:

If we can impart any high level guidance, it’s that there’s often a gap between a handful of the winningest teams and the rest. Because of that, the top seeds often do very well in the tournament, more so than the men. Here are some facts:

  1. All four No. 1 seeds have made it to the Final Four three times (compared to once for the men)
  2. Since 1982, at least one No. 1 seed has made the Final Four
  3. Four of the past six title games have been between two No. 1 seeds (compared to once for the men
  4. The lowest seed to win the national title is a No. 3 seed (compared to a No. 8 seed for the men).

That said, our models see Connecticut emerging as this year’s champion, giving Breanna Stewart and all seniors four National Championships, defeating South Carolina in a rematch of the February 8 game in Columbia, SC. The four consecutive titles would be a women’s college basketball first.

- Walter Sun, Bing Predicts Team Lead